I don’t understand the point of this, but NewsBlur seems to literally only make the unread portions of subscriptions available to Reeder – once an item is marked as read, it’s removed as if I was using POP email, which makes reading on multiple copies of Reeder impossible (again, like POP email), because I cannot return to a read item on a different device because it no longer available to be fetched.
This is not how Feedly/Old Reader/etc. integrations in Reeder work, where a capped list of the full feed’s items is fetched, even when unread, which keeps the apps in sync with each other from an end-user’s perspective (creating a more IMAP email-like experience).
I’m hoping there is some setting to change this behavior because this makes my intended use of NewsBlur impossible
Seems to me like a Reeder setting. NewsBlur is providing Reeder a list of unread stories and then Reeder downloads them from NewsBlur. But that’s fairly standard.
I’m not sure if Reeder supports the Read Stories list, but if you use the official app, it’s in there.
If a feed publishes a new item, and I read it using Reeder on the iPhone (via The Old Reader), when I later open Reeder on the Mac or the iPad it appears in that feed as a new, read item in Reeder’s “all” view.
This is how I keep my reading in sync across all of my devices in this app, and how it works with other services’ APIs like the ones I mentioned.
I’ve confirmed this just now using http://lorem-rss.herokuapp.com, where adding a new feed to The Old Reader and reading all of its items on Reeder Mac causing the next run of iPad and iPhone to add that new feed where all of its items are already marked unread.
When I do this with Newsblur, adding a new feed and reading it on Mac causing the next run of iPad and iPhone to add the new feed with absolutely 0 items in it.
Having RSS feeds for my folders was a feature I wanted to take advantage of when signing up for Newsblur, but if Newsblur’s API breaks my reading flow in Reeder this much I cannot continue using it.
I’m not sure I 100% follow. Would it be possible to post a screenshot of the behavior on the different machines? Since you went through all this trouble I’d love to resolve it for other people who may be hitting on this issue.
Honestly, no. I’ve spent several days struggling to find a service that could produce Folder RSS, and a good portion of the day already on it and verified it across my devices.
Newsblur works one way – it only exposes unread items – and Feedly, The Old Reader and others work another way. That’s fine.
This is too much effort to explain any further and I really just want to move on now and stop dealing with this.
As far as I can tell, this is a Reeder implementation decision, Reeder doesn’t seem to have any control for downloading read articles.
Don’t get me wrong, I get that Reeder behaves differently with Feedly and other backends, but, Newsblur does seem to make the articles accessible to other clients, so my guess is that (for whatever reason) Reeder doesn’t download read articles at all.
Yes, that is what I’ve been saying, it doesn’t download them at all. Reeder’s had NewsBlur support apparently since 2014 so is this a recent problem? Did the API change underneath them? Can NewsBlur escalate this with them since they’ll likely get a faster response, being the third-party they claim to support?
Hi there. I’d like to bump and raise this again, and potentially move it into the problem category. What the OP said is completely correct, Reeder is currently behaving like an “POP mail client” and is only fetching brand-new articles, not syncing with NewsBlur as one would expect. I’d be happy to describe this issue as much as possible in the hopes of seeing it fixed.
I’ve just returned to feed reading again and have chosen NewsBlur over Feedbin this time around. I love the web interface on desktop, in fact I’ve gone ahead and subscribed to Premium off the back of that alone. However, the behaviour in Reeder hasn’t been as expected. The essence is this: Reeder has three main views for any given listing; Starred (irrelevant for the purposes of this discussion,) Unread and All. When selecting the All view, I should see all of my articles, read or unread, that NewsBlur knows about (the “All Site Stories” view on web,) however, after signing into NewsBlur on Reeder, I see nothing at all.
For comparison when using Reeder with Feedbin, the All view exactly mirrored what I saw on web, and the read/unread state of each article synced appropriately, much alike an IMAP mail client. I believe this is the expected behaviour.
I’ve included two screenshots below for comparison; one of my All Site Stories in the NewsBlur app, and one of the All view in Reeder. The expected behaviour is that Reeder presents every post visible in the NewsBlur app, and the read/unread state of each article syncs appropriately.
I’ll do some investigation with some other feed readers and see if this issue is common between apps, indicating an issue with the NewsBlur API, or if this is an issue with Reeder’s implementation.
NetNewsWire replicates the behaviour. NNW doesn’t have an “All Articles” view, only an unread all articles, so I’ve pulled up Daring Fireball to show a complete lack of articles. Note that the ‘unread only’ toggle in the top right is not selected.
I suggest you bring this to Silvio’s attention. This is a change Reeder needs to make, as NewsBlur already supports an IMAP-like experience using the API directly.
I also recommend using the official apps, since they support all of NewsBlur’s features, including training and statistics. And they do a great job of fetching stories and images for offline reading.
Hi Sam, thanks for the response! I have a feeling you’re right, this does seem like a client thing. Upon further investigation it seems NNW is only syncing unread articles - if I go ahead through web and mark a bunch of articles as read they sync down just fine, and then NNW remembers them. Reeder seems a bit more finicky but NNW being open-source means I might be able to fix it myself.
Love NewsBlur on Web and have certainly been playing with the training features, however at the moment I’m leaning towards having a native client on mobile; I think I prefer the more consistent design at the expense of the reduced feature set.
